Human Pose-Constrained UV Map Estimation

TL;DR

This paper introduces PC-CSE, a novel method for UV map estimation that integrates 2D human pose to improve global coherence and anatomical plausibility in UV maps, outperforming previous pixel-wise approaches.

Contribution

We propose Pose-Constrained Continuous Surface Embeddings (PC-CSE), a new approach that incorporates 2D human pose into UV map estimation for enhanced accuracy and plausibility.

Findings

Improved UV map coherence and plausibility on DensePose COCO.

Whole-body pose constraints enhance detail in hands and feet.

Pose conditioning reduces invalid mappings.

Abstract

UV map estimation is used in computer vision for detailed analysis of human posture or activity. Previous methods assign pixels to body model vertices by comparing pixel descriptors independently, without enforcing global coherence or plausibility in the UV map. We propose Pose-Constrained Continuous Surface Embeddings (PC-CSE), which integrates estimated 2D human pose into the pixel-to-vertex assignment process. The pose provides global anatomical constraints, ensuring that UV maps remain coherent while preserving local precision. Evaluation on DensePose COCO demonstrates consistent improvement, regardless of the chosen 2D human pose model. Whole-body poses offer better constraints by incorporating additional details about the hands and feet. Conditioning UV maps with human pose reduces invalid mappings and enhances anatomical plausibility. In addition, we highlight inconsistencies in the ground-truth annotations.